Anthropic Partners With SpaceX for 220,000+ NVIDIA GPU Compute Capacity
Anthropic announced a major compute infrastructure deal with SpaceX today, securing all capacity at SpaceX’s Colossus 1 data center. The deal adds over 300 megawatts of new compute — more than 220,000 NVIDIA GPUs — expected to come online within a month.
The agreement signals a broader shift: AI companies are moving beyond renting cloud GPU instances toward owning and controlling their own compute infrastructure at industrial scale.
Alongside the partnership, Anthropic introduced several product improvements effective immediately:
- Claude Code rate limits doubled for Pro, Max, Team, and seat-based Enterprise plans
- Peak hour restrictions removed for Pro and Max accounts
- API rate limits significantly raised for Claude Opus models
This is the latest in a series of massive compute investments by Anthropic over the past year, including:
- A 5 GW agreement with Amazon (nearly 1 GW by end of 2026)
- A 5 GW agreement with Google and Broadcom (coming online in 2027)
- A strategic partnership with Microsoft and Nvidia including $30 billion in Azure capacity
- A $50 billion US AI infrastructure investment with Fluidstack
Notably, Anthropic also expressed interest in partnering with SpaceX to develop “multiple gigawatts of orbital AI compute capacity” — a potential signal that AI computing may one day extend to space-based data centers.
From an Agent Economy perspective, this deal reinforces a critical thesis: compute is no longer an operating expense — it is the defining strategic asset of the AI era. Who controls compute capacity controls the next generation of model capabilities.